Facebook Helps Blind Users Experience a Whole New Part of their Newsfeeds
Facebook’s accessibility team, headed by Jeff Wieland, has been working on ways for Facebook to interpret the increasing amount of visual content on news feeds for blind users. While screen readers can be used with Facebook, the number of photos and other objects on the platform make it tough for blind or visually impaired users to get the full experience of the social networking site.
To do this, engineers like Matt King (Facebook’s first blind engineer) are using an object recognition tool, which will describe photos and other content to users. First, it reads the caption of the photo, then provides descriptive keywords.
So how exactly can Facebook learn to recognize objects? The technology is similar to its facial recognition software, which enables Facebook to recognize a person’s face in 5 seconds using “convolutional neural networking.” More simply, Facebook uses a machine learning algorithm to teach computers to indentify faces from a database of celebrity politician photos, thereby teaching it to recognize Facebook users’ faces. This concept can be extended to other objects as well.
King told TechCrunch, “This might not be 100 percent yet, but even if it’s just halfway there, the level of engagement that’s possible, the amount of enjoyment I can get — that’s like going from zero percent to at least 50 percent of what you might get. That’s a huge jump, and it’s only going to get better from here. I personally find Facebook’s willingness to invest in ways like that just really powerful and exciting, and just one more way to make connecting people with disabilities a great experience.”
Wieland also ensures that this software will be compatible with popular assistive technology devices and programs.
Facebook plans to launch this software at the end of the year.